(1973) CD-4, SQ, QS for the Recording Engineer

QuadraphonicQuad

Help Support QuadraphonicQuad:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.
I read in Billboard around the time they chose CD-4 , that they considered both SQ and QS quite seriously.

Too bad they didn't adopt QS alongside CD-4 like some of the other labels that tried various systems the likes of Project 3 and A&M....

QS matrix would have been much more accepted stateside if they had.

Oh well, what coulda and shoulda been I guess.
 
CD-4 is the only real quad format on vinyl. The only one that claimed discreet 4-channel playback. And when it works, it is really something good. When it works with a perfect LP pressing it is great.

Getting everything to work was not going to happen for everyone all the time.

But to me it (often) delivered what the other LP formats only dreamed of.

Look at those wav file screen shots of DV SACDs, and look at a good SQ decode of the same tracks - it’s pathetic what CBS was willing to accept from a beautifully mixed quad tape to an SQ LP.
 
I read in Billboard around the time they chose CD-4 , that they considered both SQ and QS quite seriously.

Too bad they didn't adopt QS alongside CD-4 like some of the other labels that tried various systems the likes of Project 3 and A&M....

QS matrix would have been much more accepted stateside if they had.

Oh well, what coulda and shoulda been I guess.
I also read that Columbia toyed with the idea of CD-4 (a CD-4 like system) but rejected it as it was far to difficult a system to make work properly. SQ was chosen for very good reasons, full left to right separation in stereo and quad playback and full audio frequency range, no artifacting caused by use of a demodulated carrier. Columbia mixed their quad with SQ encoding and decoding in mind so most releases work very well. Others such as A&M simply SQ encoded their discrete master which I admit didn't always work that well. As you can see from my avatar SQ is the camp that I've long been in. Today I likely would't choose SQ as modern methods of encoding such as Involve have advanced farther, but today we don't really need to encode anything as discrete methods abound.
I do like the way QS Surround mode wraps the stereo around the room but it should not of been called synthesized as nothing is faked, stereo is simply pulled apart. That gets me thinking that QS Surround could/should have been used for a quad encode/decode system, all stereo would already be encoded. The addition of some anti-phase signals could fill in between the back. While that would give the rear signals more encoding space than the front there would of been no compromise with stereo and full compatibility between stereo and quad.
 
I also read that Columbia toyed with the idea of CD-4 (a CD-4 like system) but rejected it as it was far to difficult a system to make work properly.

Heck yes back then everyone was trying quadraphonic cross pollination. Even Sansui had a
patent for a carrier disc system like CD-4. The goal was to combine the qualities of matrix base-band decoding with high frequency separation enhancement. Sounds just like CD-4 but I think the consumer end product would be a simple QS decoder with the carrier signals doing the enhancement.

Even Sony had developed a Variomatrix kind of ch blending to increase separation for either QS or SQ. I can't find the patent on line but perhaps someone else can: 3,786,193.

I agree with you on both your positions of SQ/QS. I bought SQ partly because I knew hardware was in development & even with basic to more advanced I would get the wide front L/R I was used to in stereo. I was amazed when I first heard QS Variomatrix play back stereo. Incredibly better than trying to squeeze something anything out of an SQ logic decoder

And there lies the irony: SQ was the most compatible encode system for producing stereo quality sound & QS was the best for playing stereo to decoded surround!.
 
Even Sony had developed a Variomatrix kind of ch blending to increase separation for either QS or SQ. I can't find the patent on line but perhaps someone else can: 3,786,193.
You might be thinking of Paramatrix (Oh! that was CBS) I think at least some of those prototype decoders were produced. Don't forget about Lynn Olson of Audionics who developed the Shadow Vector Decoder, which was essentially Variomatrix for SQ it was only dropped due to the availability of the Tate chips.
I guess Paramatrix was actually more like Tate, Shadow Vector more like Variomatrix, sometimes I confuse the two both were promised late in the game and neither came to be, pity!
 
Last edited:
Heck yes back then everyone was trying quadraphonic cross pollination. Even Sansui had a
patent for a carrier disc system like CD-4. The goal was to combine the qualities of matrix base-band decoding with high frequency separation enhancement. Sounds just like CD-4 but I think the consumer end product would be a simple QS decoder with the carrier signals doing the enhancement.

Even Sony had developed a Variomatrix kind of ch blending to increase separation for either QS or SQ. I can't find the patent on line but perhaps someone else can: 3,786,193.

I agree with you on both your positions of SQ/QS. I bought SQ partly because I knew hardware was in development & even with basic to more advanced I would get the wide front L/R I was used to in stereo. I was amazed when I first heard QS Variomatrix play back stereo. Incredibly better than trying to squeeze something anything out of an SQ logic decoder

And there lies the irony: SQ was the most compatible encode system for producing stereo quality sound & QS was the best for playing stereo to decoded surround!.
The hybrid matrix/carrier system was actually not Sansui; rather, it was Nippon Columbia ( Denon). It was marketed in Japan as the UD-4 system. There were only a handful of records released using the system, but none made it into general distribution.
 
The hybrid matrix/carrier system was actually not Sansui; rather, it was Nippon Columbia ( Denon). It was marketed in Japan as the UD-4 system. There were only a handful of records released using the system, but none made it into general distribution.
You are correct, I think that Scott was just pointing out that Sansui too was considering such a system.
 
The hybrid matrix/carrier system was actually not Sansui; rather, it was Nippon Columbia ( Denon). It was marketed in Japan as the UD-4 system. There were only a handful of records released using the system, but none made it into general distribution.

I am well aware that Denon had a (smallish) release of commercial product using UD-4. What I noted previously was Sansui had a patent granted describing their approach to a hybrid matrix/discrete record. Not Paramatrix, not Shadow Vector. Click on the link I mentioned earlier & you'll see the assignee was Sansui.

I expect Sansui had such good sucess with Variomatrix that they never went as far as Denon to produce any releases this way.
 
You might be thinking of Paramatrix (Oh! that was CBS) I think at least some of those prototype decoders were produced. Don't forget about Lynn Olson of Audionics who developed the Shadow Vector Decoder, which was essentially Variomatrix for SQ it was only dropped due to the availability of the Tate chips.
I guess Paramatrix was actually more like Tate, Shadow Vector more like Variomatrix, sometimes I confuse the two both were promised late in the game and neither came to be, pity!

Apologies to Kirk for straying a bit off topic but I think the Sony variable matrix circuit I mentioned is worth sneaking in a couple of scans.

SONY SQ 3.jpg


This circuit is for SQ. As you can see it requires phase shift at both the input that conducts basic SQ decoding and at the output to make decode output phase corrections. The real thing to look for is the variable mixing between the chs that cancels out the crosstalk. A significant departure from commercial Sony SQ designs. The patent shows a simple bipolar transistor as the gain element with the control voltage applied to the base. Probably an FET would be a better choice & if put in production they would get an even larger linear control range using an OTA such as the RCA 3080 that was popular at that time. Of course this OTA is not bi-directional so they would have to use 2 for each mixing block.

SONY QS 3.jpg

The QS version shown above is markedly simpler compared to SQ needing only phase shift correction at the output. Not shown is the direction sensing for QS is measured on diagonal chs & you can see how the mixing for cross talk cancellation is also done on diagonal chs. Both the SQ & QS models can be characterized as having 1 fixed & 2 variable coefficients per output ch.

Years ago this patent gave me the idea to do essentially the same thing on the PC. In decoding QS or using it to synth stereo to surround, I departed from Lucanu et al & used mixing to create a 2 ch file of right front and left back, also another 2 ch file being left front right back. Using AA3 center ch cancellation I applied this to each pair, inveted left front & right front for a second cancellation and then inverted the front chs again to put the polarities right. It can all be done on a script so that's easy. Then you have to copy/paste to make the correct front L/R pair & rear L/R pair. That part is tedious but it gives me better results than any other upmixing method I've tried. Including the Surround Master because this gives you pre & post decoding control over individual songs or control of individual parts of songs.
 
Years ago, I was a fan of SQ and CD-4. Mainly because most of the albums I wanted to buy were from Columbia/Epic and WEA, each using one of the two. But I think the QS ones (compared to the SQ ones) sounded more "natural." (I had a Pioneer QX-949A with full logic SQ.

Funny story -- I worked in music radio back in the day. I wanted to add Steely Dan's "Reeling in the Years" to our playlist. I owned the QS quad album, so we dubbed it off onto a cart and used it probably daily. Our station was in suburban NYC and one of the record reps we had in NY used to listen to us on the clock radio on his desk at work. He could pick up the signal in the big CBS building in Manhattan and liked our music better than other NYC radio stations. One day he called me and asked about why that song sounded strange. What version were we using? Apparently, when listening in mono (at least after broadcast) the vocals on the song were GONE. Sounded great in stereo, though!
 
Heck yes back then everyone was trying quadraphonic cross pollination. Even Sansui had a
patent for a carrier disc system like CD-4. The goal was to combine the qualities of matrix base-band decoding with high frequency separation enhancement. Sounds just like CD-4 but I think the consumer end product would be a simple QS decoder with the carrier signals doing the enhancement.

Even Sony had developed a Variomatrix kind of ch blending to increase separation for either QS or SQ. I can't find the patent on line but perhaps someone else can: 3,786,193.

I agree with you on both your positions of SQ/QS. I bought SQ partly because I knew hardware was in development & even with basic to more advanced I would get the wide front L/R I was used to in stereo. I was amazed when I first heard QS Variomatrix play back stereo. Incredibly better than trying to squeeze something anything out of an SQ logic decoder

And there lies the irony: SQ was the most compatible encode system for producing stereo quality sound & QS was the best for playing stereo to decoded surround!.


Scott , might that have been the CBS Labs proposed "USQ" system ?

If that's the one...either Larry Clifton or Jay Frank printed a few pages in their respective quad journals.
 
I also read that Columbia toyed with the idea of CD-4 (a CD-4 like system) but rejected it as it was far to difficult a system to make work properly. SQ was chosen for very good reasons, full left to right separation in stereo and quad playback and full audio frequency range, no artifacting caused by use of a demodulated carrier. Columbia mixed their quad with SQ encoding and decoding in mind so most releases work very well. Others such as A&M simply SQ encoded their discrete master which I admit didn't always work that well. As you can see from my avatar SQ is the camp that I've long been in. Today I likely would't choose SQ as modern methods of encoding such as Involve have advanced farther, but today we don't really need to encode anything as discrete methods abound.
I do like the way QS Surround mode wraps the stereo around the room but it should not of been called synthesized as nothing is faked, stereo is simply pulled apart. That gets me thinking that QS Surround could/should have been used for a quad encode/decode system, all stereo would already be encoded. The addition of some anti-phase signals could fill in between the back. While that would give the rear signals more encoding space than the front there would of been no compromise with stereo and full compatibility between stereo and quad.

Very true Ken .:)
And of course the irony with CD-4 and CBS , is that Columbia Records pressed most if not all of the WEA quadradiscs released stateside.
 
Back
Top